Presentation (CTI)

OBJECTIVES: Health economic modeling often requires early prioritization of populations, outcomes, and intervention contexts before full evidence synthesis is feasible. In large and heterogeneous evidence bases, abstract-level evidence mapping may provide a pragmatic approach to identify candidate areas for model development. This study applied abstract-level evidence mapping to systematic reviews and meta-analyses of Tai Chi interventions as a case example for prioritizing health economic modeling.
METHODS: We conducted an abstract-level scoping review of systematic reviews and meta-analyses evaluating Tai Chi interventions. PubMed and the Cochrane Database of Systematic Reviews were searched. AI-assisted evidence management in Nested Knowledge was supplemented by human review. Eligible abstracts were extracted for population context, disease area, outcome domain, and direction of effect. Evidence was mapped separately for preventive and therapeutic contexts. Therapeutic outcomes were further distinguished into patient-relevant endpoints with direct economic relevance and functional, physiological, or behavioral outcomes that may inform model pathways.
RESULTS: Searches identified 774 records, of which 753 unique abstracts were screened and 193 reviews were included. Across included reviews, 781 outcome-level evidence datapoints were identified: 491 preventive, 67 patient-relevant therapeutic, and 223 functional, physiological, or behavioral therapeutic datapoints. Preventive Tai Chi showed broadly consistent positive evidence across functional performance, balance, quality of life, and psychological outcomes. In therapeutic contexts, the most consistent patient-relevant evidence was observed in musculoskeletal, neurological, and cardiometabolic disease areas. Functional and mechanistic outcomes were more heterogeneous but helped identify plausible intermediate pathways for model structure and assumptions.
CONCLUSIONS: Abstract-level evidence mapping can support early-stage prioritization of health economic modeling by rapidly structuring large, heterogeneous evidence bases. Applied to Tai Chi, the approach identified preventive and therapeutic contexts with decision-relevant outcome signals while separating endpoints suitable for economic model outcomes from intermediate outcomes useful for model pathways. This framework may be transferable to other complex, non-pharmacological interventions.